Frame 3: Framebuildering Takes Hold

[matthoult]

Frame 2: The First Homebuildering was a sophomoric exercise. I used fancy steel, made bilaminates, and was generally more ambitious than my skill level allowed me to execute proficiently. Knowing eyes will easily see its faults and my sloppiness, but fundamentally, it is a rideable bicycle that I learned a lot from, will continue to learn from, and enjoy.

Enter Frame 3.
Frame 3 copies the geometry for the most part from Frame 2. Frame 2 fits me quite well and I love that about it. Frame 3 however, is intended to use less expensive tubing and be simpler in its finishing. I want it to be able to blend in amongst a group of Surly’s at the rail trail headworks.
Top tube, down tube, and seat tubes will still all be 28.6mm, but whereas Frame 2 used Columbus SL tubing, Frame 3 uses Founderland 9-6-9, 4130 Taiwanese tubing from CL Cycles in Montreal. Both have the same 36mm Zona head tube stock with top and bottom reinforcement rings. Building nearly the same frame geo but with different tubing, I’m curious to see what if any differences I’ll notice.

For Frame 3, I want to keep sourcing to Canadian retailers as much as possible. Most of the steel and fittings have come from CL Cycles. For the seat and chain stays, I had initially planned to use some straight gauge 4130 from Aircraft Spruce and bend it myself, but while I was figuring that out, Danielle Schon opened https://www.framesetsupply.ca/. I am excited some Deda CX chain and seat stays will be arriving soon. Frame 2 has canti’s, but Frame 3 will have an ISO disk brake mount.

Frame 2 used a lugged bottom bracket. I remember reading it was a more forgiving approach and found that to be true. For skill development and in keeping with Frame 3’s design approach, its chain stays will be fillet brazed like the rest of the frame.

I had intended for Frame 2 to accommodate 700x43c tires. It can, but its tight, and I neglected to create practical clearance around the rear brake to remove the wheel without deflating the tire. I’ll fix this problem with Frame 3, which should fit 50c’s, or 43c’s with good fender clearance.

A Soma Wolverine fork from my parts bin should fit the frame well and get me rolling, but I do plan to make a fork to match. Most of the parts will transfer from another bike: TRP hydraulic road post mount disk brakes and Sram 1x11s with bar end shifter.

A used mini mill/lathe recently arrived in my workshop and I have been taking slow steps at learning how it works and how to set it up. The initial plan had been to use it for mitering, but hand files were in reach and the front triangle came together surprisingly quickly.

The next step is deciding on cable routing. I want to keep it external and like being able to comfortably shoulder carry.

More updates to follow.

[calstar]

Very glad to see this as I’m also a very tall guy, thanks for taking the time to post your processes. I’m guessing the difference between Columbus SL tubing and Founderland 9-6-9 4130 Taiwanese tubing will be subtle if any unless the wall thicknesses are significantly different as well as the chainstay differences.

Brian

[matthoult]

Thanks!

I’m actually not super tall. Somewhere between 6’1” and 6’2”. My saddle height is usually between 800 and 820 mm depending on fit and crank length.

I really hate the look of flat top tubes with ~8cm of spacers to put the handlebars where they need to be for comfort. My bars are a couple inches below my saddle height. Why not take up the space with headtube? Sloping tubes can look great. Big front triangles make excellent homes for frame bags. Why not use ‘em?

[matthoult]

Coming along. My brazing is getting slowly better. My heat control still needs improvement, but my technique on this one is better than on previous frames.

[matthoult]

Some learning here. Please excuse the cluttered work area. It was a byproduct of troubleshooting.

My Doug Fattic frame jig does not have a way of holding chainstays. On the last frame, this was part of the reason I went with a lugged bb.

The Alex Meade stay jig I have similarly has a single clamp fixture and no connection point to secure the bb. I only have the clamp for round chainstays, so for these I adapted seat stay clamp fixtures with spacers. For layout, it worked great, but the system was not secure enough for hole saw use and the stay moved.

What ended up working was putting the stay fixture vertical and using the angle finder to ensure everything is vertical. I doubted my ability to align them and use them correctly, but tried using BikeCAD Coping templates for the chain stays. They worked quite well and I did better with them than expected.

To make sure things were square and even, I put a spare bb shell on the cope and checked the and filed adjustments until angle was 0 degrees and the gap between the fixture extrusion and bb was the same on either side. Seemed to work okay.

The hard part came after. Squaring on a spare bb shell is easy. Squaring on a complete triangle is much harder.

In the one picture, it shows how I used my work stand to position the front triangle at the correct angle.

I considered just brazing the stays to the spare bb and installing a new front triangle onto the assembly, but after I checked and rechecked all the spacing and securing, and went for it.

I need to do another alignment. Preliminary checks show some needs, but I think it will be okay. In the future, if I’m fillet brazing the bb area on a frame, I think I will start with the stays+bb assembly, then build the front triangle onto it.

[guy153]

I do ST to BB shell first, and then the chainstays. I can just fixture them up on the table and it’s easier to position things to weld around the inside without a front triangle in the way (I like to just prop things up on the welding table, I don’t use a bike stand like some people). This means my jig only has to worry about the front triangle which makes it much simpler.

Once the CS are in place you can check the HT is right between them much easier and thus make sure your alignment is perfect.

Seatstays always go in last because then you have the option of small corrections to the chainstays if necessary right until the end.

[matthoult]

I have a small group of friends that goes on a bike ride Mondays after work. On the Friday, I decided I wanted to take Frame 3 on Monday’s ride.

It was mid-day Sunday and I hadn’t started yet. I estimated it would take me around 8h to make rideable. After 10.5 h it was rideable. I was a little optimistic.

I had aspirations of using power tools to cut the seat stays at the seat tube, but opted for BikeCAD templates. Much like my intimidation-then-relief that came with using them on the chain stays, they worked great on the seat stays.

I tried a small hacksaw blade holder to help cut the seat tube slot, but it didn’t work well. I went back to saw-blade-in-bare-hand and it went okay. A few days later, I saw a post by Richard Sachs where he used two blades at the same time in order to get the thickness he was after. Here, I made two cuts. Some good room for improvement.

While the ISO brake fitting and the brake itself worked great, the fixture didn’t quite clear the seat stay. I used a 10mm axle securing ring to offset the fixture. Ditto some nuts to offset the ISO bracket. The tolerance was about 0.3mm by my digital vernier calliper. No issues.

[matthoult]

@Schonstudio shared here about her steel bar and brake hose tool to help with finishing. It is a great tool and I’ve been meaning to make my own, but haven’t gotten around to it. The fillets between the bb, seat tube, and stays is always a tricky part to file and I was looking for something similar. I wish I had gotten around to it. What I did find was a toothbrush. the neck is firm-ish and flexible. Worked way better than my finger!

I didn’t grab a picture of it, but I used the round plates below (one is from a press. the other is from an old bearing, I think. both came from a collection of scrap metal working stuff). The 3/8” 4130 tubing bent very nicely. The diameter of the press plate is about 10”.

[matthoult]

This is the bike as built up for the Monday ride I mentioned. The cables are too short, the fillets were un-filed, and there were no bridges or brake support installed, but it was rideable and it rode great!

[matthoult]

By the week after, I had filed the fillets and the frame was looking pretty well done. The last outstanding touches for the frame (or so I thought) were just a couple cable braze-ons.

The Monday rides have become a good testing ground for the frame. They’ve mostly been rail trail, but there are opportunities to try hopping curbs and the odd log and some mild climbing.

Even with 165mm cranks (I’ve always preferred short cranks), there is just a little pedal overlap. I later switched to 160mm cranks. When I build a fork to go with the frame (the John Deere green fork in the pictures is a Soma Wolverine fork with 50mm offset), I’ll look to add just a little more offset (50mm to 55mm). The head tube angle is 72 degrees. If I were to build the frame again, I’d likely go a little more slack by a degree or two. Maybe it is just the larger tires, but the handling does feel a bit more “road-ish” than I anticipated.

On the first test ride, I didn’t notice it so much, but on last night’s ride I noticed flex in a way I’d like to change. There is a popular expression “Laterally stiff, but vertically compliant.” This frame is not laterally stiff. “Laterally flexy, but vertically compliant” is perhaps a better description. I decided to use smaller diameter tubing (three main tubes are 28.6mm, 9-6-9) when I made the bike because I do prefer a good amount flex (I’m 6’1”, 220lbs). The downside of the smaller diameter tubes built into a larger frame (head tube is 242mm long) is the amount of lateral flex. This was good learning for me. I do enjoy the art and science of frame building!

Thank you to @Daniel_Y for this thread! I look forward to applying the concepts on the next frames.

Looking at one of my favourite bikes from which I took a lot of design inspiration (Matt Chester UtiliTiman(?)), I realized that the main tubes were made with rather stout, larger diameter, straight gauge tubes, but the stays MC chose were noticeably smaller in diameter than on another Ti frame I have.

If I were to build Frame 3 again, I would opt for more rigid/larger diameter main tubes and stick with the nice 1mm-0.7mm stays. The 16mm seat stays do match the chain stays nicely, but maybe that’s the next experiment: 7/16” (11mm) or 1/2” (12mm) or 9/16” (14mm) seat stays. When I laid out the chain stays, I tried to maximize the amount of narrow end, removing the bulk from the bb end. I like that.

Now, I am faced with what to do with the flex, floppy frame that fits me so nicely and is otherwise quite comfortable. I remembered seeing Rivendell’s with double top tubes, presumably to assist with a similar big frame flex issue. I suspect the frame size, tube size, and diameters used are not altogether different. I never liked the aesthetics of the diagatube of the Bombadil, but the arched tube in the Atlantis looks pretty lovely:

My frame has three bottle bosses spaced out along the underside of the top tube and there are two sets of bottle bosses on the top of the downtube. My intention was to have a frame bag made that follows the general shape of the curved tube in the Atlantis, with the bag attaching to the top tube and the most forward set of bottle bosses–something like the profile shown in D) below.

I have a 5/8” 4130 tube that bent nicely around the 10” press plate. After the weekly ride, I tried marking up the frame with some concepts for where the extra tube might attach. I’m leaning to something like Option B), or maybe Option C). Sorry for the awkward shadows on the garage door. The downtube’s front-most bottle cage bosses might no longer be usable for a bottle, but should still work for something like a cargo deck.

Option A)

Option B)

Option C)

Option D)

[Fuzzmuffin]

For aesthetics, option c!

No comment on the structural aspect though.

[guy153]

Nice build! I would be wary of A and D because it looks like you would be joining to the thin part of the DT. You really only want to be attaching things to the butted sections.

You could also just replace the DT with a fatter one. A traditional “oversize” frame has 1 1/4” DT and the 1 1/8” TT that you have. You could try that or go even fatter.

[matthoult]

Thank you. This is likely the way.

In my eagerness in having the bike finished, I hadn’t considered the tube replacement and upsizing option. Adding a support tube seemed like the faster, easier way to ‘done.’

I couldn’t tell when I was riding if the flexing was happening more at the tt-ht, dt-ht, or both. It looked like it was mostly at the tt-ht, but that could have been because that was the spot closer to my eye and so easier to notice.

If I’m keen to change the HT angle anyway, it may make sense to replace and upsize both the TT and DT.

[bushtrucker]

Yeah I agree with this. Option C easily looks the best!

[gschwell]

C definitely looks the best. My gut tells me it won’t help with frame flex, but it would be cool to find out.

[matthoult]

My options as I see them:

1. Add the 5/8 tube and see what happens. I don’t really have any objective means of comparing with vs without. I do have subjective means of comparison, but after the tube is installed, I only have memory to compare it to.
2. Replace the downtube only with either a larger diameter 31.8 or 35mm tube. It is an opportunity to fix the bottle boss spacing. I could still add the 5/8” tube, but it would mostly be for fun.
3. Replace the dt and cut the ht off from the tt and re-angle the ht at the tt to slack out a little, as noted above. This seems tricky, would reduce the already short-ish effective tt length (575mm) and expose the tt to another heat cycle.
4. Replace the tt-ht-dt assembly with new, fixing (or trying to) all of the issues (diameter/flex, HT angle, slight pedal overlap, bottle boss location and alignment) identified.

So far, I’m leaning towards option 4. Besides the funny lateral flex, I actually really like the bike and would prefer to keep riding it, if possible. With option 4, I can build the replacement assembly in my jig over time while continuing to ride the bike. Then when I’m ready, I can hack the tt and dt off, grind off the excess, and then transplant the new front assembly onto the existing rear triangle.

Thinking about my bike while looking at some other, not dissimilar, bikes for inspiration, I’m trying to decide on tube diameters for the replacement tt and dt.

I have more 9-6-9 4130 in 28.6, 31.8, and 35mm on hand, so I’d like to go with some combination of those.

My current feeling, for whatever reason, is to go with a 35mm dt and 28.6mm tt, but that might just be because it seems like it would be a fun thing to try and a combination a little out of the norm.

[Dr.Hossa]

Or…
You can start building another complete frame, and then compare the both to each other
The knowledge gain would be even greater.

-Thom

[guy153]

Option 4 sounds the best to me for the reasons given. The 35mm tube should be about 34% stiffer in bending than the 31.75, and 84% stiffer than the 28.6 (!). But torsional stiffness may be more important. The 31.75 is about 37% stiffer than the 28.6 in bending.